Insight into Structure and Assembly of the Nuclear Pore Complex by Utilizing the Genome of a Eukaryotic Thermophile

Insight into Structure and Assembly of

the Nuclear Pore Complex by Utilizing

the Genome of a Eukaryotic Thermophile

1_{Biochemie-Zentrum der Universita¨t Heidelberg, Im Neuenheimer Feld 328, Heidelberg D-69120, Germany} 2_{EMBL, Meyerhofstrasse 1, Heidelberg D-69117, Germany}

3_{These authors contributed equally to this work}

*Correspondence:[email protected](P.B.),[email protected](E.H.) DOI10.1016/j.cell.2011.06.039

SUMMARY

Despite decades of research, the structure and

assembly of the nuclear pore complex (NPC), which

is composed of

30 nucleoporins (Nups), remain

elusive. Here, we report the genome of the

thermo-philic fungus

Chaetomium thermophilum (ct) and

identify the complete repertoire of Nups therein.

The thermophilic proteins show improved properties

for structural and biochemical studies compared

to their mesophilic counterparts, and purified

ctNups enabled the reconstitution of the inner pore

ring module that spans the width of the NPC from

the anchoring membrane to the central transport

channel. This module is composed of two large

Nups, Nup192 and Nup170, which are flexibly

bridged by short linear motifs made up of linker

Nups, Nic96 and Nup53. This assembly illustrates

how Nup interactions can generate structural

plas-ticity within the NPC scaffold. Our findings therefore

demonstrate the utility of the genome of a

thermo-philic eukaryote for studying complex molecular

machines.

INTRODUCTION

The conserved nuclear pore complexes (NPCs) spanning the double nuclear membrane are the sole mediators of transport between the nucleus and the cytoplasm. The NPC is an 50 MDa structure displaying an eight-fold symmetry and con-sisting of a number of distinct substructures: the spoke-ring complex, the cytoplasmic pore filaments, the nuclear basket, and the central channel, through which nucleocytoplasmic transport occurs (Fahrenkrog and Aebi, 2003). The NPC consists of30 different proteins named nucleoporins, or Nups, that were initially identified by nuclear fractionation, with the help of anti-bodies or lectins, through genetic screens, and by affinity purifi-cation (for review, see Doye and Hurt, 1997). Subsequently, complete NPCs were isolated revealing the entire set of Nups

for both yeast and vertebrates (Cronshaw et al., 2002; Rout et al., 2000). Each Nup is present in multiple copies (8, 16, or 32) with a total of500 Nups per NPC (Alber et al., 2007). One group of Nups contains phenylalanine-glycine (FG)-rich repeat domains that line the central transport channel, allowing shut-tling transport receptors to pass through a mesh of FG nucleo-porins via transient interactions with these FG motifs (Patel et al., 2007). The other group of Nups, which is devoid of FG repeats, constitutes the NPC core structure.

The majority of Nups are part of discrete subcomplexes, which are conserved from yeast to humans and are the building blocks of the NPC (D’Angelo and Hetzer, 2008). Biochemically defined NPC subcomplexes in yeast include the following: (1) the Y-shaped Nup84 complex (Siniossoglou et al., 1996), (2) the Nsp1 complex, a module held together by coiled-coil interac-tions and with FG repeats protruding into the central pore channel (Grandi et al., 1993), (3) the Nup82 complex, which is located at the cytoplasmic face (Belgareh et al., 1998; Hurwitz et al., 1998), (4) the Nup60-Nup2 heterodimer, which is located on the nucleoplasmic side of the NPC (Dilworth et al., 2001), (5) the Nup170-Nup53/Nup59 complex (Marelli et al., 1998), which is essential for NPC assembly (Makio et al., 2009), and (6) the Ndc1 complex, an assembly of three integral membrane proteins inserted in the pore membrane (Onischenko et al., 2009).

The two largest structural Nups from yeast, Nup192 and Nup188, have not been isolated in stable subcomplexes, but substoichiometric interactions have been reported between Nic96 and either Nup188 or Nup192 (Kosova et al., 1999; Nehrbass et al., 1996). Moreover, two-hybrid screens demon-strated an interaction between Nic96 and Nup53 (Fahrenkrog et al., 2000), implying an association between Nic96 and the Nup170 subcomplex.

The large structural Nups are predicted to have either an a-helical structure (Nup192, Nup188) or a combined b-propeller/ a-solenoid fold (Nup170, Nup157) (Devos et al., 2006). However, it remains unknown whether they directly interact with each other and how they are linked to the other NPC modules. Current models predict a cage-like or fence-like NPC scaffold formed by Nups withb-propeller and/or a-solenoid folds, such as the Y-shaped Nup84 complex (Brohawn et al., 2008; Devos et al., 2004; Hsia et al., 2007). Despite these insights into NPC structure

and assembly, key questions remain. Central to this is how the building blocks of the NPC, the 30 Nups and their derived subcomplexes, make up the major substructures of the NPC scaffold. In particular, the formation of the inner pore ring (also called spoke ring or equatorial ring) by the large Nups, Nup192, Nup188, Nup170, and Nup157, remained elusive.

In this study, we report the genome of a thermophilic eukaryote, which was found to be well suited for biochemical studies of Nups. The Nups identified in the fungus Chaetomium thermophilum enabled reconstitution of the NPC module that forms the inner pore ring. This assembly, which potentially spans from the pore membrane to the FG transport channel, consists of Nup192 and Nup170, bridged by Nic96 and Nup53. Within this complex, we have identified nucleoporin interactions that depend on short and flexible motifs. Electron microscopy demonstrates that Nup192 and Nup188 exhibit an ‘‘S’’-like morphology that resembles the shapes of karyopherins, which are shuttling nuclear transport receptors. Collectively, our find-ings support the hypothesis that the large structural Nups preferentially bind short linear motifs provided by linker Nups for assembly into the NPC scaffold.

RESULTS

Two-Hybrid Analysis Identifies New Nup Interactions To identify the connectivity and physical interactions between the structural Nups in the NPC framework, we performed a two-hybrid analysis with yeast Nups. We focused on Nups pre-dicted to forma-helical, b-propeller, or a combination of the two folds because these predominant fold-types in the NPC are thought to significantly contribute to multiple Nup interac-tions. These investigations not only recapitulated some of the known Nup associations but also revealed unidentified interac-tions including that of Nic96 with either Nup192 or Nup188 (Figures 1A and 1B). Shortening of the interacting motif in Nic96 identified a relatively short sequence in the N domain (residues 101–190) that was necessary and sufficient to bind to full-length Nup192 or Nup188 (Figure 1C), or derived C domains (Nup192553–1683and Nup188561–1655), but not smaller constructs (data not shown). Sequence analysis indicated that the binding motif in Nic96 contains a predicteda helix of 60 amino acids (residues 109–164) that is conserved with a characteristic spacing of hydrophobic residues and interspersed charged amino acids (Figures 1C and 1D). In contrast, the conserved coiled-coil domain in the N-terminal part of Nic96 (residues 1–100) that was previously shown to recruit the Nsp1-Nup49-Nup57 core complex (Grandi et al., 1993) did not interact with Nup192 or Nup188 in the two-hybrid assay (Figure 1C). Thus,

a previously unknown motif in the flexible N-terminal domain of Nic96 interacts with Nup192 and Nup188.

We performed genetic analyses to demonstrate a requirement of the Nic96 helix motif for NPC structure and assembly. Deletion of a sequence from Nic96 containing the helix motif (nic96D101– 190) caused a mild growth defect at 23C, whereas deletion of the coiled-coil domain (nic96D1 100) retarded cell growth more significantly (Figure S1A available online). However, when both the coiled-coil and the adjacent helix motifs were deleted (nic96D1–190), an extremely slow growth phenotype was generated. Thus, the synergistic intramolecular overlap between these two motifs suggests that recruitment of the Nsp1 core complex via the coiled-coil sequence and binding of Nup192 or Nup188 via the adjacent helix are essential for Nic96 function. A functional overlap of the Nic96 helix with Nup192 or Nup188 was investigated by genetic analysis. Combination of viable nup192 mutants with the nic96Dhelix allele resulted in synthetic lethality (Figure S1B). In contrast, the nic96 mutant lacking the coiled-coil sequence was not genetically linked to these nup192 alleles (Figure S1B and data not shown). Combination of the viable nup188D mutant with nic96Dhelix was not lethal but resulted in a synergistically enhanced growth defect (data not shown) with a failure to efficiently target Nup192-GFP to the NPCs (Figure S1C). These data indicate that Nic96 is func-tionally linked to Nup192 and Nup188 via its N-terminal helix motif, which facilitates assembly of these large structural Nups into the NPC.

Superior Nucleoporin Properties Enabled by a Thermophilic Eukaryote

Next we sought to verify the two-hybrid interaction found between the large structural Nups and Nic96. In order to do this, we purified Nup192, Nup188, and Nic96, either directly from yeast or recombinantly from E. coli, and performed in vitro reconstitution studies. However, these studies were not conclusive due to both a modest yield and stability of the large structural Nups (data not shown). In order to improve the outcome of these studies, we performed reconstitution studies with Nups derived from a thermophilic eukaryote. It was assumed that these Nups could display improved biochemical and structural properties after isolation, as observed for thermo-stable proteins from thermophilic prokaryotes.

Among eukaryotes, only a few microorganisms have the capa-bility to grow at temperatures above 50C (Hickey and Singer, 2004). One of them is the fungus C. thermophilum, which can thrive at temperatures of up to 60C (La Touche, 1948). This ther-mophilic, filamentous ascomycete is a member of the family Chaetomiaceae, which is found in soil, dung, or rotting plants.

Figure 1. Two-Hybrid Screen Identifies New Interactions between Yeast Nucleoporins

(A) Two-hybrid interactions between yeast Nups (indicated by black arrows). Nups with predicteda-helical (red) or b-propeller (green) folds. Nup53/Nup59 (blue). (B) Two-hybrid interaction between Nic96 and Nup188 or Nup192, and between Nup53 and Nic96. Two-hybrid plasmids expressing the indicated G4BD (GAL4 DNA-binding domain) and G4AD (GAL4 activation domain) constructs were transformed into a yeast reporter strain and grown on the indicated plates. (C) Nup188 and Nup192 interact with a predicteda helix in the Nic96 N domain. Two-hybrid interactions between the indicated Nic96 constructs and Nup188/ Nup192 (right panel).

(D) Sequence alignment, using ClustalW2 and Jalview, of Nic96/Nup93 orthologs from Saccharomyces cerevisiae (S.c.), Pichia stipitis (P.s.), Chaetomium thermophilum (C.t.), Chaetomium globosum (C.g.), Neurospora crassa (N.c.), Homo sapiens (H.s.), Xenopus laevis (X.l.), and Drosophila melanogaster (D.m.). See alsoFigure S1.

It contributes significantly to the breakdown of cellulose with a high-temperature decompostation phase (>55C), which only allows thermophilic organisms including C. thermophilum to grow. This suggests that the proteome of C. thermophilum is more stably folded than that of its mesophilic relatives. Indeed, analysis of a few enzymes (e.g., xylanases) cloned from C. thermophilum indicated thermostability at up to 60C ( Hakuli-nen et al., 2003). Remarkably, C. thermophilum is most closely related to the mesophilic mold fungus Chaetomium globosum (optimal growth temperature of 24C), a frequent indoor contaminant generating mycotoxins and encountered as a caus-ative agent of skin and nail infections in humans.

To gain insight into the morphology of C. thermophilum, we performed light and electron microscopy of cells grown at 50C. This analysis revealed that this thermophile is a myce-lium-forming fungus with multinucleate hyphae that are perfo-rated by septa as observed for other filamentous ascomycetes such as Neurospora crassa or Aspergillus nidulans (Figure S2). C. thermophilum exhibits a nucleus and a distinct nucleolus surrounded by the double nuclear membrane with interspersed NPCs. Sporadically, a spindle pole body associated with the nuclear membrane or a dividing nucleus can be seen, which points to a closed mitosis (seeDiscussion). Further conspicuous structures in the thin-sectioned cells are mitochondria with numerous christae, ER membranes, vesicles, and multivesicular bodies proving that C. thermophilum is a typical eukaryote with distinct compartments and organellar structures known from mesophilic eukaryotes.

Encouraged by our consideration to have access to thermo-stable nucleoporins, we sought to sequence the genome of C. thermophilum. We used the isolate of C. thermophilum derived from fermenting straw by La Touche in 1948, grown at 50C, to extract DNA (Figure S2A). The genome of C. thermophilum was sequenced using the Roche 454 FLX and XLR platforms (Margulies et al., 2005; Ronaghi et al., 1996) and a 24-fold coverage assembly was obtained that separated into 20 scaffolds (only 4 more than expected from the 8 chromo-somes), implying a high assembly quality (Table S1). Based on abundance differences of reads, we were also able to separate and completely assemble the mitochondrial genome (127 kb containing 15 protein-coding and 28 tRNA genes) and conclude that per nuclear genome 40 mitochondrial genomes exist. In the nuclear genome of 28.3 Mb, we predicted 243 tRNA genes and found the rRNA operon at 1500-fold coverage, but currently it cannot be assembled as part of the scaffolds. Based on 21,141 independently sequenced cDNAs sequences, the total coverage of the nuclear genome was estimated to be 96%. Using a set of 4,011 single-copy orthologs, derived for the subphylum of Pezizomycotina, the gene coverage was in close agreement to this figure at 98.6%. In total, we identified 7,227 protein-coding genes (Table S1) by ab initio gene predic-tion with AUGUSTUS (Stanke and Waack, 2003), of which 6,862 could be mapped to orthologous groups using the eggNOG pipeline (Muller et al., 2010). The genome and pro-teome sequence of C. thermophilum is available at NCBI GenBank or athttp://ct.bork.embl.de, where it can be subjected to various queries, e.g., for the C. thermophilum genes (by name, position, or PFAM domains, etc.), or where BLAST can be run

with a mesophilic protein of interest against the predicted C. thermophilum proteome. For 4,230 (73%) out of 5,797 S. cerevisiae proteins, a homolog exists in C. thermophilum that can be used for homology modeling. Similarly, there are homologs in C. thermophilum for 10,191 (44%) out of 22,937 human proteins including many families that cannot be found in prokaryotes.

Searches in the C. thermophilum genome demonstrated that all of the Nups, which are conserved from yeast to humans, are present in this eukaryotic thermophile (Table S2 and Figure S3). As gene order is only conserved within close relatives to C. thermophilum (data not shown), synteny could not be used to confirm orthology. Orthology could be confirmed by triangles of reciprocal best BLAST (Altschul et al., 1997) hits as implemented in eggNOG v2 (Muller et al., 2010) for 20 ctNups; the others were confirmed by using the ctNup as query for psiBLAST and retrieving the original Nup. Central to this study, we identified and subsequently analyzed the C. thermophilum (ct) orthologs of the yeast Nup192 (196.9 kDa), Nup188 (203.7 kDa), Nup170 (156.2 kDa), Nic96 (121.3 kDa), and Nup53 (46.9 kDa). Multiple sequence align-ments with derived average distance trees indicate that the ctNups are most closely related to the annotated orthologs of C. globosum, followed by Podospora anserina, N. crassa, and A. nidulans, which are all members of the Pezizomycotina clade. Importantly, the ctNups are related, albeit more distantly, to their well-studied mesophilic orthologs from Saccharomyces cerevisiae, Schizosaccharomyces pombe, and Homo sapiens (Figure S4).

The genes encoding these thermophilic Nups were cloned from genomic DNA or, when necessary, from cDNA (see Exper-imental Procedures) and heterologously expressed in yeast. This approach was specifically chosen because the behavior of the large structural ctNups (>150 kDa) in the eukaryotic expression system was superior to that of ctNup expression in E. coli (data not shown). Analysis of affinity-purified thermophilic Nups by SDS-PAGE and Coomassie staining revealed that the large structural Nups, ctNup192, ctNup188, and ctNup170, could be purified in significantly higher amounts than the corre-sponding yeast orthologs with little degradation or contamina-tion (Figure 2A).

The isolated ctNups exhibited an increased thermosolubility when compared to their yeast counterparts. Strikingly, purified ctNup192 and ctNup170 remained soluble at temperatures of up to 57C, which is close to the temperature at which this ther-mophilic fungus ceases growth (Figure 2B). In contrast, purified yeast Nup192 and Nup170 precipitated at 36C (Figure 2B and data not shown). Moreover, the large structural Nups eluted from the gel filtration column predominantly as monomeric proteins (Figure 2C). Thus, heterologous gene expression in yeast yielded monomeric ctNups that could be purified in high quantities (approximately 3 mg from a 10 l culture) and with increased thermostability.

In Vitro Reconstitution of Nup Interactions with Nups fromC. thermophilum

Next, we used the purified large Nups from C. thermophilum for in vitro reconstitution. Initially, we focused on the interactions

between Nic96 and Nup192 or Nup188 that were identified in the two-hybrid screen. ProtA-tagged ctNic96 (for sequence see Fig-ure S5A) was affinity-purified on IgG-Sepharose and incubated with purified ctNup192 or ctNup188 in the presence of a competing lysate. As shown by SDS-PAGE and Coomassie staining, ctNic96 requires the conserved amino-terminal helix motif identified in the yeast system to bind to either ctNup192 or ctNup188 (Figure 3A) yielding stoichiometric heterodimers (Figure 3B). Moreover, the purified ctNic96 helix alone was able to mediate binding to ctNup192 or ctNup188 (Figure S5B). For this interaction, the C-terminal domain of

Figure 2. Purification and Stability of C. thermophilum Nups

(A) Affinity purification of ctNup170, ctNup188, and ctNup192, in comparison to yeast scNup192. Lysate (lanes 1, 4, 7, 10), unbound proteins (2, 5, 8,11) and TEV-eluates (3, 6, 9, 12) were analyzed by SDS-PAGE and Coomassie staining. Affinity-purified Nups (*). TEV protease (#). Marker (M).

(B) Thermosolubility (in vitro aggregation) test for yeast and ctNups. Affinity-purified ctNup170, ctNup192, and yeast scNup192 were incubated at the indicated temperatures for 1 hr, before supernatant (S) and pellet (P) fractions obtained by centrifugation were analyzed by SDS-PAGE and Coomassie staining.

(C) Gel-filtration chromatography of the indicated purified ctNups on Superdex 200. Input and fractions were analyzed by SDS-PAGE and Coomassie staining (lower graph).

See alsoFigure S2,Figure S3,Figure S4,Table S1, and

Table S2.

ctNup192 (Nup192549–1756) was sufficient, although binding was reduced compared to the full-length protein. These data suggest that the folded Nup192 or Nup188 proteins, rather than a short sequence, interact with the helix motif of Nic96. Notably, conserved hydrophobic amino acids in the ctNic96 helix are important for binding, as their mutation to alanines reduced binding to ctNup192 (Figure S5C). Similarly, mutation of the orthologous conserved hydrophobic residues in the yeast Nic96 helix significantly diminished the two-hybrid interac-tion with yeast Nup192 or Nup188 (Figure S5C), suggesting that these mutations have a similar effect in yeast. Thus, a conserved helix motif in the Nic96 N domain mediates the formation of a complex with Nup192 or Nup188, possibly by binding to a hydrophobic patch on the sole-noid surface of these large Nups.

To clarify whether Nic96, Nup192, and Nup188 can form a heterotrimer, ProtA-ctNup188 immobilized on IgG beads was first incubated with 10-fold molar excess of ctNup192, but no binding was observed ( Fig-ure 3C). Notably, when ProtA-ctNup188 was incubated with both ctNup192 and ctNic96, binding of Nic96 to Nup188 was strongly decreased with recovery of both Nup192 and Nic96 in the unbound fraction ( Fig-ure 3C). Altogether these data indicate that stoichiometric Nup192-Nic96 and Nup188-Nic96 heterodimers but no tertiary complex can be formed in vitro. Apparently, Nup192 and Nup188 compete for the same binding site (i.e., thea helix motif) on Nic96.

Encouraged by the finding that interactions involving large ctNups can be efficiently reconstituted in vitro, we extended our studies to other ctNups. In particular, we focused on interactions between ctNup53 (C. thermophilum has only one

Nup53/Nup59 ortholog) and ctNup192, ctNup170, and ctNic96 (seeFigures 1A and 1B). We performed in vitro binding assays, essentially as outlined above using the purified ctNups present in a competitive lysate. Using this method, we were able to recon-stitute heterodimeric complexes between (1) ctNup53 and

ctNup192 (Figure 4A, lane 2), (2) ctNup53 and ctNic96 (lane 3), and (3) ctNup53 and ctNup170 (lane 4). ctNup53 also interacted with ctNup188 but less strongly than with ctNup192 (Figure S6A). To investigate which domains/motifs in Nup53 mediate these different interactions, we generated a series of Nup53 deletion

Figure 3. An_{a Helix Motif in the Nic96 N Domain Is Necessary and Sufficient for Binding to Nup192 or Nup188}

(A) ProtA-TEV-ctNic96 containing thea helix motif (for sequence seeFigure S5A) or lacking it (ctNic96Dhelix) was immobilized on IgG-Sepharose before addition of purified full-length ctNup192, ctNup188, ctNup192C, or ctNup192N mixed with an E. coli lysate (Input, +L). After binding and washing, it was eluted with TEV and the indicated eluates were analyzed by SDS-PAGE and Coomassie staining (Bound). The band (#) corresponds to endogenous yeast Nup192 coenriching with ProtA-ctNic96 during affinity purification.

(B) Stoichiometric binding of ctNic96 to either ctNup192 or ctNup188. Immobilized ProtA-TEV-ctNic96 was incubated with increasing amounts of ctNup192 or ctNup188. It was eluted by acidic pH and the indicated eluates were analyzed by SDS-PAGE and Coomassie staining.

(C) ProtA-ctNup188 immobilized on IgG-Sepharose was incubated with increasing amounts of ctNup192 or ctNic96, either separated or mixed. Input (lanes 1 and 2), bound material (eluted by acidic pH, lanes 3–7), and flowthrough fractions (lanes 8–11) were analyzed by SDS-PAGE and Coomassie staining. Mixing of ProtA-Nup188 and Nup192 revealed that SDS-PAGE can separate both bands (lane 12).

constructs. In addition to an amphipathic a helix at the C terminus that binds the pore membrane protein Ndc1 or can directly insert into membranes (Marelli et al., 2001; Onischenko et al., 2009), Nup53 contains a conserved RRM-like (RNA recog-nition motif) domain (Handa et al., 2006). Thus, ctNup53 was divided into an N fragment (residues 1–88), RRM domain (133– 252), C fragment (residues 329–389), and C-terminal amphi-pathic a helix (390–427) (Figures 5A and 5B). Following this domain classification, different ctNup53 constructs were gener-ated and tested for interaction with other Nups. ctNic96 and ctNup192 bound efficiently to the N fragment of ctNup53 ( Fig-ure 5C, lanes 1–6), whereas ctNup170 was recruited by the C fragment (Figure 5C, lanes 7–12, 15–16). In contrast, neither the RRM-like domain nor the C-terminal amphipathica helix were required for these interactions in vitro. Further shortening of the N domain identified a sequence between residues 30 to 70, which was sufficient for binding to ctNup192 or ctNic96, albeit with reduced affinity (data not shown). Whether a Phe-Gly (FG) peptide present in the Nup53 N domain (Figure 5B) is directly involved in binding to Nup192 or Nic96 awaits further structural analysis.

Finally, we investigated whether the C-terminal domain of ctNic96 (ctNic96Dhelix) and ctNup192 can simultaneously bind to the N fragment of ctNup53 or whether these interactions are

mutually exclusive. ctNic96Dhelix was still capable of interacting with the Nup53-N motif (we used the larger Nup53-N+RRM construct that can be better determined on the SDS-polyacryl-amide gel), whereas interaction with ctNup192 was abolished (Figure S5D; see alsoFigure 3A). Similarly, when ctNic96Dhelix was incubated with both ctNup53-N+RRM and ctNup192, only Nup53 but no Nup192 was coenriched (Figure S5D). Because Nup192 per se is competent to interact with the Nup53-N motif (see Figure 5C), these data suggest that Nup192 and Nic96 compete for the same binding site within the Nup53 N domain. In Vitro Assembly of the Inner Nuclear Pore

Ring Complex

In light of our observation that ctNic96 and ctNup53 have short motifs involved in multiple contacts to the large structural Nups, we extended the binding studies and tested whether more complicated assemblies can be reconstituted between the various structural ctNups. In order to do this, we immobilized GST-ctNup53 on beads and sequentially added the respective purified ctNups. In this way, it was possible to reconstitute heterotrimeric complexes between ctNup192, ctNup170, and GST-ctNup53 (Figure 4A, lane 5); ctNic96, ctNup170, and GST-ctNup53 (Figure 4A, lane 7); and ctNup192, ctNic96, and GST-ctNup53 (Figure 4A, lane 9). Both the Nup192-Nup170-Nup53

Figure 4. Reconstitution of the Tetrameric Nup192-Nup170-Nic96-Nup53 Complex withC. thermophilum Nucleoporins

(A) GST-ctNup53 was immobilized on GSH beads before addition of purified ctNic96, ctNup170, and ctNup192 (input, lanes 11–13), in the presence of competitive E. coli lysate. After binding and washing, it was eluted with SDS-sample buffer and eluates were analyzed by SDS-PAGE and Coomassie staining (lanes 1–10).

(B) Reconstitution of the ctNup192-Nup170-Nic96-Nup53 module by using immobilized ProtA-ctNup192 and addition of purified ctNup170, ctNic96, and ctNup53DC1 or ctNup53DC2 in the presence of an E. coli lysate. Bound material was eluted by TEV and analyzed by SDS-PAGE and Coomassie staining. See alsoFigure S6.

Figure 5. Nup53 Is a Multivalent Binding Protein Interacting with Nup192, Nic96, and Nup170

(A) Domain organization of ctNup53. Below the schematic drawing are ctNup53 deletion constructs used for the in vitro binding studies.

(B) Amino acid sequence alignment of Nup53 orthologs from Chaetomium thermophilum (C.t.), Neurospora crassa (N.c.), Magnaporthe oryzae (M.o.), Aspergillus clavatus (A.c.), and Aspergillus niger (A.n.).

and Nic96-Nup170-Nup53 complexes were efficiently formed even in the presence of excess ctNup170. This is consistent with the observation that ctNup170 binds to the C domain, whereas ctNup192 or ctNic96 bind to the N domain of ctNup53 (see above). Finally, a heterotetrameric complex between ctNup192, ctNup170, ctNic96, and GST-ctNup53, in which all subunits were present in stoichiometric amounts due to the cooperative binding of ctNup192 and ctNic96, could be efficiently reconstituted (Figure 4A, lane 10). Similarly, this heter-otetrameric complex could be formed in a different way by first immobilizing ProtA-Nup192 on IgG-Sepharose followed by reciprocal addition of the other purified components (Figure 4B). Recruitment of ctNup170 to this module was solely dependent on the C motif in Nup53. This was apparent because a ctNup53 construct lacking this motif (ctNup53DC2) was associated with the absence of ctNup170 from the module, whereas a ctNup53 construct carrying this motif (ctNup53DC1) was associated with inclusion of ctNup170 in the module (Figure 5C, lanes 17– 20 andFigure 4B).

Finally, a related tetrameric complex could be assembled in which ctNup192 was replaced by ctNup188. However, this tetramer contained slightly reduced amounts of ctNup188 (Figures S6B and S6C), which is explained by the observation that ctNup53 has lower affinity to ctNup188 than to ctNup192 (seeFigure S6A). In conclusion, we have reconstituted an NPC module composed of the large solenoid Nup192 (or Nup188) together with Nup170, Nic96, and Nup53 and defined the motifs/domains by which these four Nups interact. Most of the Nups that are part of this module are implicated in formation of the inner nuclear pore ring (seeDiscussion).

Structural Insight into the Components of the Inner Pore Ring Module

To gain structural insight into the components of the inner pore ring, we used electron microscopy (EM) to analyze the structure of ctNup192, ctNup188, and ctNup170. Overview pictures of the negatively stained proteins revealed a homogeneous distri-bution of particles. Accordingly, ctNup170 exhibits a crescent-shaped structure with one bulbous end (15 nm length, 6 nm width), which is similar to yeast Nup170 (Figure 6A, see Flem-ming et al., 2009).

Markedly, the EM structures of ctNup192 and ctNup188 significantly differ from ctNup170. Although not similar in their primary sequence, ctNup192 and ctNup188 exhibit comparable S-like structures (18 nm length, 10 nm width) with the two semi-circles twisted by approximately 90 (Figure 6B, Figure S7A). Labeling of ctNup192 with a prominent EM label (DID-Dyn2) (Flemming et al., 2010) revealed that the N and C termini are located at the respective tips of the S-shaped molecule ( Fig-ure 6C). When comparing the unusual S-like morphology of the a-helically predicted Nup192 molecule to known structures, we noticed that each of the two half circles of Nup192 resembles,

in both shape and curvature, the karyopherin transport receptors (e.g., exportin-t or Crm1) that share a superhelical architecture of tandem HEAT repeats (Figure 6D,Movie S1).

Moreover, to mark the contact site of Nic96 on the surface of the S-shaped Nup192 molecule, the structural DID-Dyn2 label was first attached to the ctNic96 helix followed by incubation with purified ctNup192. Biochemical analysis indicated an efficient binding of the labeled Nic96 helix to Nup192, and EM revealed that the attached structural label stuck out from the S-shaped Nup192 molecule (Figure S7B). These data are consistent with the two-hybrid and biochemical data, which indi-cated that the Nic96 helix binds optimally to full-length Nup192, but that reduced binding to the C domain (Nup192549–1756) was still possible.

We also investigated the trimeric Nup192-Nic96-Nup53 and tetrameric Nup192-Nup170-Nic96-Nup53 complexes by nega-tive-stain EM and selected in both cases approximately 10,000 single particles of each for image analysis. In the case of the trimer, class averages were obtained revealing the typical S-shaped Nup192 molecule but with a diagonally overlaid elongated structure. This density could correspond to Nic96, known to be an 8 nm long rod-like molecule (Figure S7C) (seeSchrader et al., 2008). Nup53 consists of a globular RRM domain, approximately 3 nm in size (Handa et al., 2006), with attached short and flexible N and C domains. However, the precise position of Nup53 in this EM structure is unclear. In the case of the negatively stained tetrameric complex, a larger variety of class averages was obtained, which we attribute to structural variability of the inner pore ring module due to a more flexible attachment of Nup170 to the Nup192-Nic96-Nup53 core complex (data not shown).

In conclusion, EM analysis revealed the morphology of the Nups constituting the inner pore ring module and uncovered an unexpected structural similarity of a-helically predicted Nup192 and Nup188 to karyopherins. This latter observation supports speculation about the evolution of Nups and shuttling transport receptors that are members of the karyopherin super-family (seeDiscussion).

DISCUSSION

In this study, we exploit a thermophilic eukaryote for biochemical and structural investigations of large nucleoporins that have previously proven to be difficult to handle when derived from several mesophilic organisms including yeast and humans. Due to the superior biophysical and biochemical properties of C. thermophilum proteins, we were able to reconstitute and assemble a conserved module of the nuclear pore complex. This study raises the possibility that this thermophile could serve as a model organism for numerous other eukaryotic complexes that, to date, have been difficult to handle biochemically. By providing a public genome and proteome of a thermophilic

(C) Distinct motifs in ctNup53 bind ctNup192, ctNic96, and ctNup170. Binding studies were performed on GSH beads with immobilized full-length GST-ctNup53 and the indicated deletion constructs to which the indicated purified proteins in the presence of E. coli lysate were added. ctArx1 (50 kDa) served as negative control.

eukaryote, we hope to have created a rich resource for structural biology and biotechnology, which should be valuable for a broad range of applications including structural proteomics and biotechnological developments.

Annotation of the ensemble of NPC proteins from C. thermophilum indicated that all of the known conserved Nups from various mesophilic organisms are present in this thermophile (Table S2, Figure S3, and Figure S4). In some respects, the C. thermophilum nucleoporins resemble those of vertebrates better than the model organism yeast does. For example, in contrast to yeast, we identified orthologs of Nup37 and Elys

Figure 6. EM Analysis of _{ctNup170 and} ctNup192

(A and B) Electron micrographs of purified ctNup170 (A) and ctNup192 (B). Overviews (left) and class averages (top row) of particles deter-mined by multivariate statistical analysis agree with reprojections of the final 3D model (middle) and surface representations of the equivalent orientations (bottom).

(C) Labeling of ctNup192 at the C terminus (left) and N terminus (right) with the DID-Dyn2 label. The arrow indicates the emergence of the label from the ctNup192 molecule.

(D) Three-dimensional reconstruction of ctNup192 with indicated N and C termini (left) compared to the atomic structures and calculated density maps (30 A˚ resolution) of exportin-t (red, PDB: 3IBV) and CRM1 (blue, PDB: 3GB8).

See alsoFigure S7,Movie S1, andTable S3.

in the C. thermophilum genome, as has been shown for A. nidulans, which is also a member of the Pezizomycotina clade. Therefore, the corresponding ctNup84 complex appears to be more related to the vertebrate Nup107-Nup160 complex than to the yeast Nup84 complex (D’Angelo and Hetzer, 2008). On the other hand, C. thermophilum is closely related to N. crassa and A. nidulans (see above), which both exhibit a closed mitosis with no breakdown of the nuclear envelope. Thus, we predict that the thermophile also divides by closed mitosis. However, it remains unknown whether the NPC partially disassembles during mitosis as observed for A. nidulans (De Souza et al., 2004), or whether it stays intact as described for N. crassa (Roca et al., 2010). Development of a transformation system combined with homologous inte-gration of a GFP reporter at different ctNUP loci could enable us to differentiate between these possibilities.

C. thermophilum also contains ortho-logs of the nuclear pore membrane span-ning Nups including Ndc1, Pom152, and Pom34, which may be linked to the structural Nups of the inner pore ring (i.e., Nup170, Nup53; see below). In the case of the Nups that constitute the nuclear basket, we have identified a clear C. thermophilum ortholog of Mlp1 (Tpr1 in higher eukary-otes); however, Nup1 (Nup153 in vertebrates), Nup2, and Nup60 could not be identified based on homology searches. In general, the Nups forming the nuclear basket are poorly conserved in evolution, making it difficult to assign corresponding orthologs. However, two putative ctNups, ctNup56 and ctNup152, which contain FG repeats and predicted Ran-binding domains (Table S2andFigure S3), are candidates for nuclear basket proteins.

Our studies revealed that the large ctNups have excellent biochemical and structural properties, allowing the reconstitu-tion of one of the long sought-after key modules of the NPC, the inner pore ring complex. This achievement was possible because two-hybrid and genetic analyses in yeast could be combined with in vitro assembly of the orthologous Nups from the thermophile. The reconstituted NPC module has the poten-tial to span the entire width of the NPC, from the anchoring membrane to the central FG repeat-containing transport channel (Figure 7). This conclusion is supported by several findings. First, one subunit of the module, Nup53, can directly bind via its C-terminal amphipathica helix to the pore membrane involving the integral membrane protein Ndc1 (Onischenko et al., 2009). Second, Nup170 in yeast is functionally linked to another pore membrane protein Pom152 (Marelli et al., 2001; Tcheperegine et al., 1999), suggesting that a Nup170-Pom152 interaction further contributes to pore membrane anchorage of the inner pore ring module. Third, another subunit, Nic96, can recruit the Nsp1-Nup49-Nup57 complex that is part of the central FG trans-port channel. Fourth, the estimated size of the Nup192-Nup170-Nic96-Nup53 tetramer (approximately 20–25 nm) is consistent with such a complete bridging, taking into account the known dimensions of the central FG transport channel (approximately 38 nm) and the overall NPC diameter (approximately 98 nm) (Alber et al., 2007).

This investigation provided insight into the mechanism of how the large structural proteins Nup192, Nup188, and Nup170/ Nup157 interact with Nic96 and Nup53. In the past, hints as to how these Nups could be organized in subcomplexes came from studies not only in yeast (see Introduction) but also in higher eukaryotes, which revealed that vertebrate Nup53 can interact with a number of Nups including Nup93 (yNic96), Nup155 (yNup170), Nup205 (yNup192), and Ndc1 ( Hawryluk-Gara et al., 2005). However, it remained unclear whether Nup205, Nup188, and Nup93 are part of a subcomplex (Miller et al., 2000) or are assembled in separate modules (Theerthagiri et al., 2010). Our studies have demonstrated that heterodimeric

Figure 7. Model of the Inner Pore Ring Complex in the NPC

Schematic drawing of the NPC in the nuclear envelope. The inner pore ring (green) is composed of Nup192 and Nup170, bridged by short and flexible motifs provided by linker Nups, Nup53 and Nic96. Nup53 is connected to the pore membrane protein Ndc1 and Nup170 to Pom152. Nic96 binds via the helix motif to Nup192 and via the coiled-coil domain (cc) to the Nsp1 complex, from which FG repeat filaments protrude into the central pore channel.

complexes, Nup192-Nic96 and Nup188-Nic96, can be assembled, but a supercomplex between Nup192, Nup188, and Nic96 is not formed. Thus, binding of the Nic96a helix motif to the two large solenoid Nups is mutually exclusive. This finding supports the data from the Antonin lab, which revealed that vertebrate Nup93 is part of two distinct subcomplexes, interacting either with Nup205 or Nup188 (Theerthagiri et al., 2010).

Moreover, we discovered that the large structural Nups of the inner pore ring do not directly interact but are flexibly bridged by short linear motifs provided by linker Nups. This indicates a previ-ously unrecognized mode of Nup interaction within the NPC. Previously described Nup-Nup contacts involve coiled-coil inter-actions (e.g., within the Nsp1/p62 complex) and interinter-actions between folded a-helical domains (e.g., between Nup84 and Nup145C; Nagy et al., 2009) or between foldeda-helical and b-propeller domains (e.g., between Sec13 and Nup145; Bro-hawn and Schwartz, 2009). Here, we found that in the case of Nic96, the interacting motif is a conserved predicted a helix protruding from the flexible Nic96 N domain and docking onto the solenoid surface of either Nup192 or Nup188, an interaction that appears to be mutually exclusive. Similarly, in the case of Nup53, short linear motifs present in the flexible N and C domains make direct contact with all three inner pore ring Nups, Nup192, Nic96, and Nup170, thereby stabilizing the entire module by multiple ‘‘crosslinks.’’ Such a mode of interaction could allow for structural flexibility and variability within the inner pore ring module that directly adjoins to a layer of Nups that line the central transport channel (Figure 7). In this way, the central transport channel could dilate or shrink depending on the size of the passing transport cargo.

The excellent biophysical properties of the ctNups also facili-tated single particle EM analysis, which gave insight into the structural organization of the large solenoid Nups. When comparing the unusual S-like morphology of thea-helically pre-dicted Nup192 molecule to known structures, we noticed that each of the two half circles of Nup192 resembles, in both shape and curvature, karyopherin transport receptors (e.g., exportin-t or Crm1) (Figure 6D andMovie S1). Very little is known about the fold type of the two largest Nups, Nup188 and Nup192/ Nup205, in any eukaryote, except that they are mainlya helical and adopt a SPAH (stacked pairs of a-helices) fold (Devos et al., 2006). Indeed, bioinformatic analysis indicates that

Nup192 and Nup188 are similar to karyopherins (e.g., Crm1, exportin-t) in terms of tertiary structure (Table S3). Accordingly, it is possible that Nups and karyopherins share a common evolu-tionary origin (Devos et al., 2006). One plausible scenario could be that an ancestral Nup (Nup192-like) may have lost its ability to stably assemble into the NPC, e.g., by altering a binding groove for a linear ‘‘Nup motif’’ on its solenoid surface so that a ‘‘NES motif’’ could be adopted. Thus, the binding mode of a SPAH Nup to a linear motif of a linker Nup could be similar to the mechanism of how karyopherins interact with cargo.

In conclusion, our analyses using nucleoporins from a thermo-phile identified a central complex of the NPC and offer support to the possibility of an evolutionary relationship between the gate and the transporters, the NPC and the karyopherins.

EXPERIMENTAL PROCEDURES

A detailed description of the methods employed in this study is provided in the

Extended Experimental Procedures.

Strains, Media, and Plasmids

Plasmids generated by standard procedures are listed inTable S4. Saccharo-myces cerevisiae strains used in this study are listed inTable S5.

Cloning of Nup Genes from_{C. thermophilum}

C. thermophilum La Touche var. thermophilum was obtained from DSMZ, Braunschweig, Germany. Genomic DNA and total RNA were isolated accord-ing to standard procedures. cDNA was synthesized usaccord-ing Superscript-III Reverse Transcriptase (Invitrogen) and an Oligo d(T) primer according to the company’s instructions. cDNA was purified with the QIAquick PCR purification Kit (QIAGEN). C. thermophilum Nup-encoding genes were cloned from either cDNA or genomic DNA using standard procedures. The ctNup ORFs were in-serted into yeast or E. coli expression plasmids.

Genome Sequencing and Analysis

DNA extraction was performed as previously described (Al-Samarrai and Schmid, 2000). The genome was sequenced using 454 GS FLX and 454 GS XLR technologies (Margulies et al., 2005; Ronaghi et al., 1996). These were assembled using Celera assembler (Miller et al., 2008) v5.3 resulting in 20 scaf-folds with 243 coverage. Gene prediction was done by AUGUSTUS (Stanke and Waack, 2003) and manually for the mitochondrial genome. Protein-coding genes were annotated with SMART domains (Letunic et al., 2009).

Two-Hybrid Analysis

For yeast two-hybrid analysis, the yeast reporter strain PJ69-4A was cotrans-formed with the yeast Nup constructs fused to the GAL4-binding domain or GAL4 activation domain (James et al., 1996).

Affinity Purification and In Vitro Binding Assays

Affinity purification of ctNups using ProtA, GSH, or FLAG tags was performed on IgG-Sepharose, GSH beads, or FLAG-antibody beads.

Thermosolubility Test of Yeast andC. thermophilum Nups

To test for thermostability, yeast Nup170 and Nup192 and ctNup192, ctNup188, and ctNup170, expressed in yeast under the control of the consti-tutive ADHI promoter, were affinity-purified via the ProtA-tag and incubated for 1 hr at different temperatures.

Electron Microscopy and Image Processing

Negative stain EM using 2% (w/v) uranyl acetate and image processing were performed as described (Flemming et al., 2010). Spatial orientations of class averages were calculated by sinogramm correlation and combined to a three-dimensional (3D) map using the weighted back-projection algorithm.

Labeling of ProtA-ctNup192 or ProtA-ctNic96 helix with the structural Dyn2-DID marker was performed according toFlemming et al. (2010).

ACCESSION NUMBERS

The nuclear genome sequences of Chaetomium thermophilum var. thermophilum DSM 1495 reported in this paper have been deposited at DDBJ/ EMBL/GenBank under the accession number ADUW00000000. The version described in this paper is the first version, ADUW01000000. The complete mitochondrial genome has been deposited at DDBJ/EMBL/GenBank under the accession number JN007486. Moreover, the DNA and protein sequences of the ctNups described in this study have been deposited at NCBI with the accession numbers indicated inTable S2.

SUPPLEMENTAL INFORMATION

Supplemental Information includes Extended Experimental Procedures, seven figures, five tables, and one movie and can be found with this article online at

doi:10.1016/j.cell.2011.06.039.

ACKNOWLEDGMENTS

We thank Eurofins MWG Operon, in particular Dr. Georg Gradl, for a very competent cooperation in the DNA sequencing project. We thank Christian von Mering for help using the eggNOG pipeline and Sheila Lutz for initial help in setting up blast searches using the C. thermophilum genomic sequence. We thank Andrea Hellwig (Prof. Hilmar Bading, Department of Neurobiology, University of Heidelberg) for performing thin-section electron microscopy. E.H. is recipient of grants from the Deutsche Forschungsgemein-schaft (SFB 638/B2). Received: November 30, 2010 Revised: April 15, 2011 Accepted: June 24, 2011 Published: July 21, 2011 REFERENCES

Al-Samarrai, T.H., and Schmid, J. (2000). A simple method for extraction of fungal genomic DNA. Lett. Appl. Microbiol. 30, 53–56.

Alber, F., Dokudovskaya, S., Veenhoff, L.M., Zhang, W., Kipper, J., Devos, D., Suprapto, A., Karni-Schmidt, O., Williams, R., Chait, B.T., et al. (2007). The molecular architecture of the nuclear pore complex. Nature 450, 695–701. Altschul, S.F., Madden, T.L., Scha¨ffer, A.A., Zhang, J., Zhang, Z., Miller, W., and Lipman, D.J. (1997). Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Res. 25, 3389–3402. Belgareh, N., Snay-Hodge, C., Pasteau, F., Dagher, S., Cole, C.N., and Doye, V. (1998). Functional characterization of a Nup159p-containing nuclear pore subcomplex. Mol. Biol. Cell 9, 3475–3492.

Brohawn, S.G., Leksa, N.C., Spear, E.D., Rajashankar, K.R., and Schwartz, T.U. (2008). Structural evidence for common ancestry of the nuclear pore complex and vesicle coats. Science 322, 1369–1373.

Brohawn, S.G., and Schwartz, T.U. (2009). Molecular architecture of the Nup84-Nup145C-Sec13 edge element in the nuclear pore complex lattice. Nat. Struct. Mol. Biol. 16, 1173–1177.

Cronshaw, J.M., Krutchinsky, A.N., Zhang, W., Chait, B.T., and Matunis, M.J. (2002). Proteomic analysis of the mammalian nuclear pore complex. J. Cell Biol. 158, 915–927.

D’Angelo, M.A., and Hetzer, M.W. (2008). Structure, dynamics and function of nuclear pore complexes. Trends Cell Biol. 18, 456–466.

De Souza, C.P., Osmani, A.H., Hashmi, S.B., and Osmani, S.A. (2004). Partial nuclear pore complex disassembly during closed mitosis in Aspergillus nidulans. Curr. Biol. 14, 1973–1984.

Devos, D., Dokudovskaya, S., Alber, F., Williams, R., Chait, B.T., Sali, A., and Rout, M.P. (2004). Components of coated vesicles and nuclear pore complexes share a common molecular architecture. PLoS Biol. 2, e380. Devos, D., Dokudovskaya, S., Williams, R., Alber, F., Eswar, N., Chait, B.T., Rout, M.P., and Sali, A. (2006). Simple fold composition and modular architec-ture of the nuclear pore complex. Proc. Natl. Acad. Sci. USA 103, 2172–2177. Dilworth, D.J., Suprapto, A., Padovan, J.C., Chait, B.T., Wozniak, R.W., Rout, M.P., and Aitchison, J.D. (2001). Nup2p dynamically associates with the distal regions of the yeast nuclear pore complex. J. Cell Biol. 153, 1465–1478. Doye, V., and Hurt, E. (1997). From nucleoporins to nuclear pore complexes. Curr. Opin. Cell Biol. 9, 401–411.

Fahrenkrog, B., and Aebi, U. (2003). The nuclear pore complex: nucleocyto-plasmic transport and beyond. Nat. Rev. Mol. Cell Biol. 4, 757–766. Fahrenkrog, B., Hu¨bner, W., Mandinova, A., Pante´, N., Keller, W., and Aebi, U. (2000). The yeast nucleoporin Nup53p specifically interacts with Nic96p and is directly involved in nuclear protein import. Mol. Biol. Cell 11, 3885–3896. Flemming, D., Sarges, P., Stelter, P., Hellwig, A., Bo¨ttcher, B., and Hurt, E. (2009). Two structurally distinct domains of the nucleoporin Nup170 cooperate to tether a subset of nucleoporins to nuclear pores. J. Cell Biol. 185, 387–395. Flemming, D., Thierbach, K., Stelter, P., Bo¨ttcher, B., and Hurt, E. (2010). Precise mapping of subunits in multiprotein complexes by a versatile electron microscopy label. Nat. Struct. Mol. Biol. 17, 775–778.

Grandi, P., Doye, V., and Hurt, E.C. (1993). Purification of NSP1 reveals complex formation with ‘GLFG’ nucleoporins and a novel nuclear pore protein NIC96. EMBO J. 12, 3061–3071.

Hakulinen, N., Turunen, O., Ja¨nis, J., Leisola, M., and Rouvinen, J. (2003). Three-dimensional structures of thermophilic beta-1,4-xylanases from Chae-tomium thermophilum and Nonomuraea flexuosa. Comparison of twelve xylanases in relation to their thermal stability. Eur. J. Biochem. 270, 1399– 1412.

Handa, N., Kukimoto-Niino, M., Akasaka, R., Kishishita, S., Murayama, K., Terada, T., Inoue, M., Kigawa, T., Kose, S., Imamoto, N., et al. (2006). The crystal structure of mouse Nup35 reveals atypical RNP motifs and novel homodimerization of the RRM domain. J. Mol. Biol. 363, 114–124.

Hawryluk-Gara, L.A., Shibuya, E.K., and Wozniak, R.W. (2005). Vertebrate Nup53 interacts with the nuclear lamina and is required for the assembly of a Nup93-containing complex. Mol. Biol. Cell 16, 2382–2394.

Hickey, D.A., and Singer, G.A. (2004). Genomic and proteomic adaptations to growth at high temperature. Genome Biol. 5, 117.

Hsia, K.C., Stavropoulos, P., Blobel, G., and Hoelz, A. (2007). Architecture of a coat for the nuclear pore membrane. Cell 131, 1313–1326.

Hurwitz, M.E., Strambio-de-Castillia, C., and Blobel, G. (1998). Two yeast nuclear pore complex proteins involved in mRNA export form a cytoplasmically oriented subcomplex. Proc. Natl. Acad. Sci. USA 95, 11241–11245. James, P., Halladay, J., and Craig, E.A. (1996). Genomic libraries and a host strain designed for highly efficient two-hybrid selection in yeast. Genetics 144, 1425–1436.

Kosova, B., Pante´, N., Rollenhagen, C., and Hurt, E. (1999). Nup192p is a conserved nucleoporin with a preferential location at the inner site of the nuclear membrane. J. Biol. Chem. 274, 22646–22651.

La Touche, C.J. (1948). A Chaetomium-like thermophile fungus. Nature 161, 320.

Letunic, I., Doerks, T., and Bork, P. (2009). SMART 6: recent updates and new developments. Nucleic Acids Res. 37(Database issue), D229–D232. Makio, T., Stanton, L.H., Lin, C.C., Goldfarb, D.S., Weis, K., and Wozniak, R.W. (2009). The nucleoporins Nup170p and Nup157p are essential for nuclear pore complex assembly. J. Cell Biol. 185, 459–473.

Marelli, M., Aitchison, J.D., and Wozniak, R.W. (1998). Specific binding of the karyopherin Kap121p to a subunit of the nuclear pore complex containing Nup53p, Nup59p, and Nup170p. J. Cell Biol. 143, 1813–1830.

Marelli, M., Lusk, C.P., Chan, H., Aitchison, J.D., and Wozniak, R.W. (2001). A link between the synthesis of nucleoporins and the biogenesis of the nuclear envelope. J. Cell Biol. 153, 709–724.

Margulies, M., Egholm, M., Altman, W.E., Attiya, S., Bader, J.S., Bemben, L.A., Berka, J., Braverman, M.S., Chen, Y.J., Chen, Z., et al. (2005). Genome sequencing in microfabricated high-density picolitre reactors. Nature 437, 376–380.

Miller, B.R., Powers, M., Park, M., Fischer, W., and Forbes, D.J. (2000). Iden-tification of a new vertebrate nucleoporin, Nup188, with the use of a novel organelle trap assay. Mol. Biol. Cell 11, 3381–3396.

Miller, J.R., Delcher, A.L., Koren, S., Venter, E., Walenz, B.P., Brownley, A., Johnson, J., Li, K., Mobarry, C., and Sutton, G. (2008). Aggressive assembly of pyrosequencing reads with mates. Bioinformatics 24, 2818–2824. Muller, J., Szklarczyk, D., Julien, P., Letunic, I., Roth, A., Kuhn, M., Powell, S., von Mering, C., Doerks, T., Jensen, L.J., and Bork, P. (2010). eggNOG v2.0: extending the evolutionary genealogy of genes with enhanced non-supervised orthologous groups, species and functional annotations. Nucleic Acids Res. 38(Database issue), D190–D195.

Nagy, V., Hsia, K.C., Debler, E.W., Kampmann, M., Davenport, A.M., Blobel, G., and Hoelz, A. (2009). Structure of a trimeric nucleoporin complex reveals alternate oligomerization states. Proc. Natl. Acad. Sci. USA 106, 17693– 17698.

Nehrbass, U., Rout, M.P., Maguire, S., Blobel, G., and Wozniak, R.W. (1996). The yeast nucleoporin Nup188p interacts genetically and physically with the core structures of the nuclear pore complex. J. Cell Biol. 133, 1153–1162. Onischenko, E., Stanton, L.H., Madrid, A.S., Kieselbach, T., and Weis, K. (2009). Role of the Ndc1 interaction network in yeast nuclear pore complex assembly and maintenance. J. Cell Biol. 185, 475–491.

Patel, S.S., Belmont, B.J., Sante, J.M., and Rexach, M.F. (2007). Natively unfolded nucleoporins gate protein diffusion across the nuclear pore complex. Cell 129, 83–96.

Roca, M.G., Kuo, H.C., Lichius, A., Freitag, M., and Read, N.D. (2010). Nuclear dynamics, mitosis, and the cytoskeleton during the early stages of colony initi-ation in Neurospora crassa. Eukaryot. Cell 9, 1171–1183.

Ronaghi, M., Karamohamed, S., Pettersson, B., Uhle´n, M., and Nyre´n, P. (1996). Real-time DNA sequencing using detection of pyrophosphate release. Anal. Biochem. 242, 84–89.

Rout, M.P., Aitchison, J.D., Suprapto, A., Hjertaas, K., Zhao, Y., and Chait, B.T. (2000). The yeast nuclear pore complex: composition, architecture, and trans-port mechanism. J. Cell Biol. 148, 635–651.

Schrader, N., Stelter, P., Flemming, D., Kunze, R., Hurt, E., and Vetter, I.R. (2008). Structural basis of the nic96 subcomplex organization in the nuclear pore channel. Mol. Cell 29, 46–55.

Siniossoglou, S., Wimmer, C., Rieger, M., Doye, V., Tekotte, H., Weise, C., Emig, S., Segref, A., and Hurt, E.C. (1996). A novel complex of nucleoporins, which includes Sec13p and a Sec13p homolog, is essential for normal nuclear pores. Cell 84, 265–275.

Stanke, M., and Waack, S. (2003). Gene prediction with a hidden Markov model and a new intron submodel. Bioinformatics 19 (Suppl 2), ii215–ii225. Tcheperegine, S.E., Marelli, M., and Wozniak, R.W. (1999). Topology and func-tional domains of the yeast pore membrane protein Pom152p. J. Biol. Chem. 274, 5252–5258.

Theerthagiri, G., Eisenhardt, N., Schwarz, H., and Antonin, W. (2010). The nucleoporin Nup188 controls passage of membrane proteins across the nuclear pore complex. J. Cell Biol. 189, 1129–1142.